Anthropic & the Claude Family

Claude Code is one product in a family built by Anthropic, and knowing the whole lineup pays off — each family member solves a different job. This lesson maps the products: what each one is for, how each is paid, and where Claude Code fits.

In this lesson:

• The company and the models behind everything
• The product family: Claude, Cowork, Claude Code, API
• Why Claude Code is the developer's pick

📌One company, one engine

Anthropic is an AI research company known for its Claude models. Every product below runs on this same model family. The current tiers, top to bottom:

• Fable — the newest and most capable tier (the first "Mythos-class" model), built for the hardest reasoning and longest autonomous work
• Opus — the powerful flagship for complex tasks
• Sonnet — the balanced workhorse; most day-to-day coding runs here
• Haiku — fastest and cheapest, for light and high-volume tasks

The products don't differ in intelligence — they differ in what that intelligence is packaged to do: chat, office work, or code. Inside Claude Code you switch models with the /model command; Module 4 covers when each tier is worth it.

📌Claude — the assistant

The product most people mean by "Claude": a conversational assistant at claude.ai in the browser, plus desktop apps (Mac/Windows) and mobile apps (iOS/Android). Writing, analysis, research, working with documents — general knowledge work, in a chat.

Billing is a subscription: a Free tier for light use, Pro at $20/month ($17/month billed annually), and Max from $100/month with 5–20× the usage for heavy daily work.

📌Claude Cowork and Claude Code — the agents

Two products move Claude from answering to doing: you delegate a task, the agent works through it, you review the result.

• Claude Cowork — non-coding work: organize folders of files, build spreadsheets, prepare reports, run recurring tasks. Lives in the Claude desktop app.
• Claude Code — the same delegation model for software: it reads your codebase, edits files, runs commands and tests. Lives in your terminal, IDE and browser.

Both are included from the Pro plan up — no separate purchase. Anthropic also ships specialized editions (Claude Design, Claude Science, Claude Security) — the same idea, packaged for specific fields.

📌Claude API — for building products

The fourth way in is the Claude Platform (API): your software calls the Claude models directly. Different purpose — building your own AI-powered products — and different billing: no subscription, you pay per million tokens processed. For example, Claude Sonnet 5 costs $3 per million input tokens and $15 per million output ($2/$10 introductory pricing through August 2026).

The distinction matters for your own projects: if you use Claude, you subscribe; if you build on Claude, you pay per token.

📌Why this course is about Claude Code

For a developer, Claude Code is the most powerful member of the family. The chat assistant can discuss code you paste into it; Claude Code works in your actual repository — reads it, changes it, runs it, tests it — using the same toolchain you use.

It's also the deepest product to master: custom instructions, skills, hooks, subagents, MCP integrations. That depth is what the two parts of this course cover — first configuring Claude Code to fit your projects, then plugging it into the wider ecosystem of ready-made tools.

One practical note: the $20/month Pro plan covers everything taught in this course.

✅Remember

• Anthropic builds the Claude models — Fable, Opus, Sonnet, Haiku — one engine behind every product
• Claude = chat assistant (browser, desktop, mobile) · Cowork = agent for office work · Claude Code = agent for software
• The API is pay-per-token, for building your own products
• Pro ($20/month) includes Claude, Cowork and Claude Code — one subscription covers this whole course